Global and regional genetic association analysis of ulcerative colitis and type 2 diabetes mellitus and causal validation analysis of two-sample two-way Mendelian randomization.

BACKGROUND

Clinical co-occurrence of UC (Ulcerative Colitis) and T2DM (Type 2 Diabetes Mellitus) is observed. The aim of this study is to investigate the potential causal relationship between Ulcerative Colitis (UC) and Type 2 Diabetes Mellitus (T2DM) using LDSC and LAVA analysis, followed by genetic verification through TSMR, providing insights for clinical prevention and treatment.

METHODS

Genetic loci closely related to T2DM were extracted as instrumental variables from the GWAS database, with UC as the outcome variable, involving European populations. The UC data included 27,432 samples and 8,050,003 SNPs, while the T2DM data comprised 406,831 samples and 11,914,699 SNPs. LDSC and LAVA were used for quantifying genetic correlation at both global (genome-wide) and local (genomic regions) levels. MR analysis was conducted using IVW, MR-Egger regression, Weighted median, and Weighted mode, assessing the causal relationship between UC and diabetes with OR values and 95% CI. Heterogeneity and pleiotropy were tested using Egger-intercept, MR-PRESSO, and sensitivity analysis through the "leave-one-out" method and Cochran Q test. Subsequently, a reverse MR operation was conducted using UC as the exposure data and T2DM as the outcome data for validation.

RESULTS

Univariable and bivariable LDSC calculated the genetic correlation and potential sample overlap between T2DM and UC, resulting in rg = -0.0518, se = 0.0562, = 0.3569 with no significant genetic association found for paired traits. LAVA analysis identified 9 regions with local genetic correlation, with 6negative and 3 positive associations, indicating a negative correlation between T2DM and UC. MR analysis, with T2DM as the exposure and UC as the outcome, involved 34 SNPs as instrumental variables. The OR values and 95% CI from IVW, MR-Egger, Weighted median, and Weighted mode were 0.917 (0.8480.992), 0.949 (0.8001.125), 0.881 (0.7790.996), 0.834(0.7230.962) respectively, with IVW -value < 0.05, suggesting a negative causal relationship between T2DM and UC. MR-Egger regression showed an intercept of -0.004 with a standard error of 0.009, = 0.666, and MR-PRESSO Global Test -value > 0.05, indicating no pleiotropy and no outliers detected. Heterogeneity tests showed no heterogeneity, and the "leave-one-out" sensitivity analysis results were stable. With UC as the exposure and T2DM as the outcome, 32 SNPs were detected, but no clear causal association was found.

CONCLUSION

There is a causal relationship between T2DM and UC, where T2DM reduces the risk of UC, while no significant causal relationship was observed from UC to T2DM.